21694
J. Phys. Chem. B 2006, 110, 21694-21700
pH-Dependent Phase Behavior of Carbohydrate-Based Gemini Surfactants. Effect of the Length of the Hydrophobic Spacer Jaap E. Klijn, Marc C. A. Stuart, Marco Scarzello, Anno Wagenaar, and Jan B. F. N. Engberts* Physical Organic Chemistry Unit, Stratingh Institute, UniVersity of Groningen, 9747 AG Groningen, The Netherlands ReceiVed: July 26, 2006; In Final Form: September 4, 2006
The phase behavior of a series of carbohydrate-based gemini surfactants with varying spacer lengths was studied using static and dynamic light scattering between pH 2 and 12. Cryo-electron microscopy pictures provide evidence for the different morphologies present in solution. The spacer length of the gemini surfactants was varied from two to 12 methylene units. At near neutral pH, spherical vesicles were obtained for gemini surfactants with a spacer shorter than 10 methylene units, whereas nonspherical vesicles were obtained for spacer lengths of 10 and 12. Upon decreasing the pH, the vesicles underwent transitions toward worm-like micelles and spherical micelles for a spacer length of six and larger, whereas for shorter spacers, these transitions are not observed. For the shortest spacer at low pH, perforated vesicles are observed, and vesicles built from the gemini surfactant with a spacer of four methylene units only underwent a transition toward worm-like micelles. Upon increasing the pH to slightly basic values, flocculation followed by redispersion upon charge reversal was observed up to a spacer length of eight methylene units. The redispersal is explained by hydroxideion binding to the uncharged vesicular surface. By contrast, vesicles formed from the gemini surfactants with 10 and 12 methylene units only undergo a transition toward inverted phases. The observations can be understood in terms of the packing parameter.
Introduction When two classical surfactants are covalently linked through a spacer, gemini surfactants are obtained.1 Recently, gemini surfactants with two tertiary amine functionalities and carrying (reduced) carbohydrate units bound to the headgroups (Scheme 1A) have been synthesized as nonviral gene-delivering vehicles.2,3 In the absence of DNA, they display an amazingly rich phase behavior, not only upon variation of the headgroup and spacer4,5 but also in mixtures with other vesicle- and micelleforming amphiphiles.6 Depending on the pH, both, one, or neither of the two nitrogen atoms is protonated. At a high degree of protonation, aggregates with high curvatures are formed, whereas at low degrees of protonation, aggregates with lower curvatures such as vesicles and oil droplets are observed.4-6 Interestingly, upon charge neutralization of the cationic vesicles and of oil droplets, hydroxide ions are specifically adsorbed at the aggregate/water interface at slightly basic pH. For a typical carbohydrate-based gemini surfactant, unilamellar vesicles can be formed at near neutral pH.7 Morphological transitions are observed by static and dynamic light scattering and confirmed by cryo-electron microscopy. Decreasing the pH leads to the gradual formation of worm-like micelles, initially coexisting with the remaining vesicles. The transition is accompanied by a strong decrease in the intensity of scattered light and a decrease of the maximum in the observed size distribution. At a specific pH (typically around pH 6), the conversion is completed as evidenced by a low but constant intensity of scattered light. Further lowering the pH by one unit results in the transformation of worm-like micelles into spherical micelles, * Corresponding author. E-mail:
[email protected].
SCHEME 1: General Structure of Previously Studied Gemini Surfactants (n ) 0, 1)a
a (A) R can be a methyl, a reduced monosaccharide, or a methylterminated oligo(ethylene oxide). In panel B, the gemini surfactants presented here are shown. The value of m can be 2 (1), 4 (2), 6 (3), 8 (4), 10 (5), or 12 (6).
and the intensity of scattered light finally drops to a value close to the background signal. Upon going in the opposite direction (i.e., from neutral pH going up), the electrostatic repulsion between vesicles is gradually lowered, ultimately leading to flocculation of the vesicles. However, at an even more basic pH, they redisperse with the same vesicular size distribution as before flocculation due to specific adsorption of hydroxide ions. These effects have been discussed in detail in previous papers.4,5 Here, we present the results for a series of gemini surfactants of which the size of the hydrophobic spacer is varied between two and 12 carbon atoms (Scheme 1B; m ) 2-12). It is known that for gemini surfactants with bisquaternary ammonium headgroups, the cmc increases with increasing spacer length but decreases again above a spacer length of about six carbon atoms.8,9 However, when the surface area of the surfactant at
10.1021/jp064774z CCC: $33.50 © 2006 American Chemical Society Published on Web 10/11/2006
Behavior of Carbohydrate-Based Gemini Surfactants
J. Phys. Chem. B, Vol. 110, No. 43, 2006 21695
Figure 1. Relative scattered intensity (A) and size distributions (B) of solutions containing gemini amphiphile 1 (b), 2 (0), and 3 (9) as a function of pH. For clarity, in panel B only the maximum in the size distribution is given. Between approximately pH 8 and 9, data could not be obtained due to phase separation.
the air-water interface is considered, the decrease starts at about 10 carbon atoms in the spacer. This suggests that the spacer begins to contribute to the hydrophobic volume only at more than six carbon atoms but that back-folding of the spacer into the hydrophobic interior of the aggregate does not occur until the spacer contains 10 carbon atoms. For the systems under study here, these observations are important since the length of the spacer will affect both the pKa values of the nitrogen atoms as well as the average packing parameter. The packing parameter P, developed by Israelachvili and Ninham,10 is helpful in explaining the observed phase behavior. This parameter provides a prediction of the morphology of the aggregate by approximating its surface curvature.
P)
V a0lc
(1)
In this equation, V is the volume of the hydrophobic part of the molecule, a0 is the mean cross-sectional headgroup surface area, and lc is the length of the extended all-trans alkyl tail. For a small packing parameter (1), there is a negative curvature leading to inverted structures (e.g., hexagonal, cubic, etc.). Experimental Procedures Materials. Gemini surfactants were synthesized as described previously.7 Detailed synthetic procedures will be reported in a forthcoming paper.11 Hepes, Mes (4-morpholinoethanesulfonic acid), APS (3-amino-1-propanesulfonic acid), and taurine (2aminoethanesulfonic acid) were purchased from Sigma and used as received. Sample Preparation. Solutions of gemini surfactant in chloroform were dried under a stream of nitrogen. Traces of residual solvent were removed under vacuum. To obtain small unilamellar vesicles (SUVs) with a narrow and reproducible size distribution, the lipid films were hydrated at room temperature in bidistilled water containing 5 mM each of the buffer substances Hepes, Mes, and taurine and 10 mM of APS at a pH close to the pH of flocculation of the gemini surfactant being
studied (unless stated otherwise), vortexed for several minutes, briefly tip-sonicated ( 8.
Starting from m ) 6, decreasing the length of the spacer results in a strong decrease of the pH at which spherical micelles transform into worm-like micelles to values below 2. If it is assumed that the transition from spherical to worm-like micelles requires the same value of the packing parameter for all gemini surfactants, the decrease in spacer length, leading to an increase in the packing parameter at the same protonation state, alone cannot explain this large effect. Considering that this transition is associated with a high degree of protonation of the gemini surfactants suggests that the pKa values of at least one of the nitrogen atoms is significantly decreased. Upon increasing the spacer length from m ) 6 to m ) 12, the pH of the transition from spherical micelles to worm-like micelles is only weakly affected. The intermolecular N-N distance of the gemini surfactants at the aggregate surface (Scheme 2) is determined by a balance of repulsive forces between headgroups of neighboring gemini surfactants in the aggregate and attractive van der Waals forces. This distance adapts itself as a response to external stimuli (e.g., changes in the degree of protonation). For spacers longer than the intermolecular N-N distance, the intramolecular N-N
Behavior of Carbohydrate-Based Gemini Surfactants
Figure 8. Plot of the pH of the transitions as a function of the length of the spacer. The 2 indicates the pH at which worm-like micelle formation from vesicles starts, O indicates the pH at which the formation of spherical micelles from worm-like micelles starts, 9 indicates the highest pH at which only spherical micelles are observed, 3 indicates the pH of flocculation, [ indicates the pH of redispersion, and × indicates a transition specified in the text.
J. Phys. Chem. B, Vol. 110, No. 43, 2006 21699 distance is about equal to the intramolecular N-N distance in 2 and 3 since the pH of the transition drops strongly going from m ) 6 to m ) 4. Similarly, the pH of the transition from worm-like micelles to vesicles strongly drops below a spacer length of m ) 4. One could argue that the packing parameter of the gemini surfactants forming perforated vesicles is close to the packing parameter in worm-like micelles,14-17 but nevertheless, it is justified to conclude that in worm-like micelles, the intermolecular nitrogen distance is about the same as the intramolecular nitrogen atom distance in 2. This is in agreement with the lower degree of protonation in worm-like micelles and hence weaker electrostatic repulsion between the nitrogen atoms of neighboring gemini surfactants. Upon going from m ) 2 to m ) 8, the pH of flocculation increases only by about 1 pH unit. Flocculation is related to removal of the last protons from the surface leading to a low positive zeta potential.7,23 Hence, neighboring charges are too far apart to be affected by one another. As a consequence, changes in the pH of flocculation are more likely due to (small) changes in the hydration of the interface. The pH of redispersion, coming from adsorption of hydroxide ions to the interface,7,23 is affected in a similar way as the pH of flocculation. However, for spacers with m > 8, no redispersion is observed due to the formation of inverted phases. Conclusion
Figure 9. Trends in the packing parameter at the same protonation state as a function of the spacer length for changes in hydrophobic volume and surface area. The values on the y-axis depend on the protonation state but range from 0 to values close to 1.21 The magnitude of the slopes of the lines is arbitrarily chosen. The location of the breaks in the plot are based on literature observations (see text).8,9
SCHEME 2: Schematic Presentation of the Dependence of the pKa Values on the Length of the Spacer for a Long (Left) and a Short (Right) Spacera
a
The arrows denote the shortest N-N distance.
distance is determined in a similar way, and therefore, it will be at least as long as the intermolecular N-N distance. At high degrees of protonation, neighboring surfactant molecules then largely determine the pKa value of the second nitrogen atom. Hence, this largely explains the constant pH for the transition from spherical micelles to worm-like micelles for gemini surfactants with m g 6. On the contrary, when the spacer is shorter than the intermolecular N-N distance, the intramolecular distance will then mainly determine the pKa value of the second nitrogen atom. Consequently, the pKa value of the second nitrogen atom will decrease. This leads to the conclusion that at the surface of spherical micelles,22 the intermolecular N-N
A detailed study of the pH-dependent phase behavior of a series of carbohydrate-based gemini surfactants with varying lengths of the hydrophobic spacers (m) has been performed. The pH has been varied between pH 2 and 12, leading to different degrees of protonation of the amine functionalities in the surfactant headgroup. The series contains gemini surfactants with a hydrophobic spacer varying in length between two and 12 methylene units. The transitions from vesicles toward aggregates of higher curvatures, such as the transitions from vesicles to worm-like micelles and from worm-like to spherical micelles, are strongly affected for spacers of m < 6. If m is 2 and 4, no spherical micelles are observed in the pH range studied, and the same holds for worm-like micelles for a spacer of m ) 2. Instead, for m ) 2, perforated vesicles are observed, which can be seen as branched and multi-connected worm-like micelles. The decrease in the pH of these transitions is the result of an increase of the packing parameter at the same protonation state upon decreasing values of m. As a consequence, a higher degree of protonation is required to observe the structural transitions. However, for m ) 2 and 4, the intramolecular N-N distance becomes smaller than the intermolecular N-N distance leading to hampering of double protonation. Therefore, it can be concluded that the intermolecular N-N distance in spherical and worm-like micelles is comparable to the intramolecular N-N distance of gemini surfactants with m ) 6 and 4, respectively. Almost complete deprotonation of the amine groups at the vesicular surface occurs at pH 8 and is rather independent of the length of the spacer (only 1 pH unit increase going from m ) 2 to 8). Adsorption of hydroxide ions to the vesicular surface starting at pH 9 follows the same trend. Interestingly, for spacers with m ) 10 and 12, inverted phases are formed, consistent with back-folding of the spacer. Acknowledgment. J.E.K. acknowledges the National Research School Combination Catalysis for financial support.
21700 J. Phys. Chem. B, Vol. 110, No. 43, 2006 Supporting Information Available: Static and dynamic light scattering data of 1-4 and 6. Static and dynamic light scattering data of 2 as a function of time. Cryo-electron microscopy pictures of 1, 2, and 5. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Zana, R. AdV. Colloid Interface Sci. 2002, 97, 205-253. (2) Kirby, A. J.; Camilleri, P.; Engberts, J. B. F. N.; Feiters, M. C.; Nolte, R. J. M.; So¨derman, O.; Bergsma, M.; Bell, P. C.; Fielden, M. L.; Rodriguez, C. L. G.; Guedat, P.; Kremer, A.; McGregor, C.; Perrin, C.; Ronsin, G.; van Eijk, M. C. P. Angew. Chem., Int. Ed. 2003, 42, 14481457. (3) Bell, P. C.; Bergsma, M.; Dolbnya, I. P.; Bras, W.; Stuart, M. C. A.; Rowan, A. E.; Feiters, M. C.; Engberts, J. B. F. N. J. Am. Chem. Soc. 2003, 125, 1551-1558. (4) Johnsson, M.; Engberts, J. B. F. N. J. Phys. Org. Chem. 2004, 17, 934-944. (5) Klijn, J. E.; Scarzello, M.; Stuart, M. C. A.; Wagenaar, A.; Engberts, J. B. F. N. Org. Biomol. Chem. 2006, 4, 3569-3570. (6) Scarzello, M.; Klijn, J. E.; Wagenaar, A.; Stuart, M. C. A.; Hulst, R.; Engberts, J. B. F. N. Langmuir 2006, 22, 2558-2568. (7) Johnsson, M.; Wagenaar, A.; Stuart, M. C. A.; Engberts, J. B. F. N. Langmuir 2003, 19, 4609-4618. (8) Alami, E.; Beinert, G.; Marie, P.; Zana, R. Langmuir 1993, 9, 1465-1467. (9) Wettig, S. D.; Verrall, R. E. J. Colloid Interface Sci. 2001, 235, 310-316.
Klijn et al. (10) Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. J. Chem. Soc., Faraday Trans. 2 1976, 72, 1525-1568. (11) Wagenaar, A.; Engberts, J. B. F. N., manuscript in preparation. (12) Edwards, K.; Almgren, M.; Bellare, J.; Brown, W. Langmuir 1989, 5, 473-478. (13) Almgren, M.; Edwards, K.; Karlsson, G. Colloids Surf., A 2000, 174, 3-21. (14) Edwards, K.; Gustafsson, J.; Almgren, M.; Karlsson, G. J. Colloid Interface Sci. 1993, 161, 299-309. (15) Silvander, M.; Karlsson, G.; Edwards, K. J. Colloid Interface Sci. 1996, 179, 104-113. (16) Gustafsson, J.; Ora¨dd, G.; Lindblom, G.; Olsson, U.; Almgren, M. Langmuir 1997, 13, 852-860. (17) Gustafsson, J.; Ora¨dd, G.; Almgren, M. Langmuir 1997, 13, 69566963. (18) Size distributions were obtained by fitting a Gaussian function to intensity as a function of the logarithm of the particle diameter. The width of the size distribution is then the width at half-height. (19) Siegel, D. P.; Green, W. J.; Talmon, Y. Biophys. J. 1994, 66, 402414. (20) Gustafsson, J.; Nylander, T.; Almgren, M.; Ljusberg-Wahren, H. J. Colloid Interface Sci. 1999, 211, 326-335. (21) At high protonation states, mainly structures with high curvatures (low packing parameters) are formed. Hence, the y-axis ranges typically from 0 to about 0.5. At low protonation states, the structures with low curvatures (high packing parameters) are formed. Hence, the y-axis ranges typically from 0 to about 1.3. (22) Close to the pH of the transition to worm-like micelles. (23) Johnsson, M.; Wagenaar, A.; Engberts, J. B. F. N. J. Am. Chem. Soc. 2003, 125, 757-760.
